The present invention relates to radio communication systems, radio communication devices, and radio relay devices. For example, the present invention can be applied to a radio communication system that decides a radio communication device serving as a parent to participate in a radio network.
In the case where each of radio communication devices (hereinafter, also referred to as “nodes”) cannot directly communicate with a management device that manages the nodes in a radio network system including the nodes and the management device, the radio network system may adopt a multi-hop network in which instead of a node that cannot perform direct communication, another node relays the communication. Such a multi-hop network enables a node at a long distance to communicate with the management device.
In the case where a node (hereinafter, also referred to as “participation node”) that newly participates in the multi-hop network cannot directly communicate with the management device when the participation node participates in (joins) the network, the participation node selects a node (hereinafter, also referred to as “parent node”) that relays a network participation request message instead of the participation node, from among nodes that have already joined the network. In order to select such a parent node, the participation node broadcasts a neighbor node discovery message for discovering nodes (hereinafter, also referred to as “neighbor nodes”) around the participation node.
A neighbor node that has received the message returns a response message to the participation node. At this time, if a transmission timing of the response message from the neighbor node is the same as another neighbor node, message collision happens and the participation node cannot receive the response message. Therefore, after receiving the message, the neighbor node waits for random time and then transmits the response message to avoid the collision.
There are various methods for selecting an optimum parent node. For example, one of the methods is to select a node having the highest received signal strength (RSSI value) of a received message. The participation node can decides the optimum parent node by collecting response messages from all neighbor nodes and selecting the parent node from among all the neighbor nodes. However, in order to collect the response messages from all the neighbor nodes, the participation node has to wait for maximum waiting time that may be selected in view of random waiting time for the response messages. Here, since more neighbor nodes results in higher probability of response message collision, it is necessary to limit the collision probability to a certain level or lower. Therefore, the maximum value of the random waiting time have to be set to be a little longer in view of the case of a large number of neighbor nodes.
As described above, since the participation node takes a lot of time to select the parent node, time it takes to join the network is unfortunately prolonged.
Technologies described in JP 2009-302694A and JP 2011-523830T are technologies of methods for selecting an optimum parent node in a multi-hop network. According to the technologies described in JP 2009-302694A and JP 2011-523830T, a participation node receives information on a neighbor node (the number of current child nodes, the number of hops to management device, round-trip time (RTT) to reach management device, and the like) from the neighbor node, and selects a neighbor node that seems to be optimum as a parent node.
In addition, according to JP 2011-523830T, each node calculates a packet loss rate or signal-to-noise ratio (SN ratio) from a communication history with neighbor nodes, and dynamically changes the optimum parent node in accordance with such information even after participated in the network.